Figma’s CEO: Why AI makes design, craft, and quality the new moat for startups | Dylan Field

What happens when a design tool refuses to be mediocre?

Imagine a product that quietly remade the way teams imagine software — not by being the cheapest, or the fastest to market, but by caring about craft. That is the through-line of Dylan Field's story: a belief that great design is now the leverage point for winning software. I left this conversation with a renewed sense that taste, speed, and a willingness to be playful can be strategic, not just decorative.

From a near-sale to renewed velocity

There’s an odd, energizing honesty to the tale of the Adobe deal that didn’t close. Rather than retreat into disappointment, Dylan and his team held frequent check-ins, reset expectations, and created a surprisingly humane option: Detach, a temporary severance program that let a small percentage step away. That fork in the road became a reset — and a recommitment to moving faster and building outward from their core strengths.

Why that reset mattered

What really struck me was how leadership treated clarity as oxygen. Regular communication, frank updates, and clear choices (stay and sprint or step away) stopped the rumor mill and focused energy. It wasn’t motivation by pep talk — it was management by context.

Fun as a strategic differentiator

Here’s the surprising idea: make your product fun. FigJam started as a reasonable product decision but felt soul-less. The team chose to make collaboration playful — and that choice paid off. I’m still smiling at the mental image of a design team arguing over whether “fun” counts as a defensible product strategy — then proving it does.

When play unlocks utility

FigJam’s playful features weren’t frivolous. They were precisely aimed at changing human behavior in remote brainstorms: draw out ideas, coax quiet people to speak, and create a low-friction space for contribution. Context matters — what’s welcome in a whiteboard can be distracting inside a precision design tool — and FigJam respected that boundary.

MakerWeek, muscle memory, and product taste

Figma’s internal rituals deserve attention. MakerWeek is more than a hackathon; it’s cultural engineering. Short sprints of creative experiments surface product ideas and teach the company how to ship new surfaces quickly. Watching the demos during MakerWeek apparently still fires people up — and several major features began that way. If you want innovation that actually lands, make prototyping a company ritual.

Growing taste is intentional work

Dylan’s definition of taste is refreshingly practical: build a point of view, expand your repertoire, and practice judgment. Taste isn’t mysticism — it’s a loop of exposure, reflection, and curation. The more you study adjacent disciplines, the better your design instincts will become.

Follow the workflow to expand

Figma’s expansion strategy reads like a practical map: trace a user’s workflow and fill the gaps. Whiteboarding led to FigJam; slides, sites, dev tooling, and now Make — an AI-backed prototyping surface — all came from asking, "What’s next after a user finishes this step?" It’s not a spreadsheet exercise about total addressable market. It’s product empathy turned into product taxonomy.

Time-to-value beats feature lists

The company obsessively measures how quickly a user reaches a “wow” moment. Shortening that path improves activation and retention. Dylan was candid: don’t wait to perfect every pixel before you release. Launch a useful starting point, then iterate. But do not ship an experience that blocks users from seeing the product’s core value.

Figma Make: prototypes, production, and quality guardrails

Make is Figma’s bet on AI for prototyping and internal apps. Its promise is provocative: allow anyone to sketch options, produce working prototypes, and accelerate the design conversation. But Dylan is careful: AI is a starting point, not a finish line. The product’s success depends on visual quality, design system consistency, and rigorous QA. He pulled a launch when outputs risked looking dangerously derivative — a reminder that trust and quality matter more than speed for AI features.

What to watch next

• Interoperability: Make should be a starting point with smooth round trips into Figma Design and developer workflows.
• Design systems in AI: Consistency prevents good ideas from being ignored because they look sloppy.
• Rigorous evals: QA and evaluation processes are now a non-negotiable product discipline.

Final thought: design as competitive moat

If you want one sentence to carry from this conversation, it’s this: great design is no longer optional. In an era when many products can be built quickly, craft and taste are the differentiators that win customers and loyalty. I left the talk convinced that leaders who prioritize clarity, remove blockers, and institutionalize playful experimentation will build products that don’t just exist — they resonate.

Reflective note: I walked away thinking about my own products and meetings differently — what small ritual, single reset, or playful nudge could make a real difference?